Add MiniMax speech TTS to the Media Center extension

[octo-patch]

Reason: Add MiniMax speech TTS to the Media Center extension.

Changes:

• Add an optional minimax provider to the media speech_synthesize action; the default provider behavior is unchanged.
• Call the MiniMax Text to Audio v2 endpoint for the global and China regional endpoints with model, text, voice_setting, audio_setting, pronunciation_dict, and output_format request fields.
• Validate MiniMax responses through base_resp.status_code and expose data.audio / data.status in the action output.
• Register MINIMAX_API_KEY as a Media authorization service so the key stays server-side.

Checks:

• bun test apps/server/src/extensions/media-center.test.ts
• bun test apps/server/src/extensions/storage.test.ts apps/server/src/extensions/media-center.test.ts
• bun build apps/server/src/extensions/media-center.ts

[Devin-AXIS]

Hi! Thanks so much for your continued support and for contributing this PR to the project.

We really appreciate the work you've put into improving the Media Center. We're also actively planning to expand our TTS capabilities and related media extensions, so your contribution is very timely.

We'd love to explore a deeper collaboration and discuss ideas around the future roadmap. If you're interested, feel free to reach out to me at [devin@ipollo.ai](mailto:devin@ipollo.ai). It would be great to connect and see how we can work together.

Thanks again for your support and contribution—we're looking forward to chatting with you!

[octo-patch]

Thanks for the context. I updated the MiniMax speech TTS adapter, resolved the latest main-branch conflict, and added request and response validation coverage. I ran the Media Center test file (41 tests) and the server typecheck.